Professional Documents
Culture Documents
08953451
08953451
The oil and gas industry has witnessed a revolution in fluids technology
for hydraulic fracturing. Starting in the mid 1980s, focused research led
Today, new additives and fluids are extending these capabilities and
For help in preparation of this article, thanks to Vic 1. For more on hydraulic fracturing:
Joyce, Ken Nolte and Jon Mitchell, Dowell, Tulsa, Waters AB: “Hydraulic Fracturing—What Is It?,” Jour-
Oklahoma, USA; Terry Greene, Dowell, Montrouge, nal of Petroleum Technology 33 (August 1981): 1416.
France; Richard Marcinew, Dowell, Calgary, Alberta, Veatch RW Jr, Moschovidis ZA and Fast CR: “An
Canada; and Mike Morris, Dowell, Corpus Christi, Overview of Hydraulic Fracturing,” in Gidley JL,
Texas, USA. Holditch SA, Nierode DE and Veatch RW Jr (eds):
In this article, HIGHSHEAR, CleanFLOW and PropNET Recent Advances in Hydraulic Fracturing, Monograph
are marks of Schlumberger. 12. Richardson, Texas, USA: Society of Petroleum
Engineers (1989): 1-38.
34 Oilfield Review
Pump trucks
Continuous fluid mixer Proppant blender
nFracturing operation. Modern treatments rely on a variety of process-controlled mixing, blending and pumping equipment
and computer monitoring and recording systems, which permit real-time decision making.
Autumn 1995 35
Unfractured, radial flow Low-permeability, linear flow
The Rock, the Mechanics and the Fluid where the tip is ideally the last area to be permanent logs of job results. Engineers
Historically, fracturing has been applied pri- packed. tracking the progress of the treatment use
marily to low-permeability—0.1 to 10 md— Why the different approach? The answer is graphic displays that plot actual pumping
formations with the goal of producing nar- found in the relationship between fracture parameters against design values to facilitate
row, conductive flow paths that penetrate length and the permeability contrast real-time decision making. Production simu-
deep into the reservoir. These less restrictive between the fracture and the formation. lators compare treatment results with expec-
linear conduits replace radial flow regimes Where the contrast is large, as for low-per- tations, providing valuable feedback for
and yield a several-fold production increase meability reservoirs, longer fractures provide design of the next job.
(above ). For large-scale treatments, as many proportionally greater productivity. Where At the heart of this complex process is the
as 40 pieces of specialized equipment, with the contrast is small, as in high-permeability fracturing fluid.3 The fluid, usually water-
a crew of 50 or more, are required to mix, formations, greater fracture length provides based, is thickened with high molecular
blend and pump the fluid at more than 50 minimal improvement. Fracture conductivity weight polymers, such as guar or hydrox-
barrels per minute (bbl/min) [8 m 3/min]. is, however, directly related to fracture ypropyl guar. It must be chemically stable
Pumping may last eight hours with width. Using short—about 100-ft [30-m]— and sufficiently viscous to suspend the prop-
1,000,000 gal of fluid and 2,000,000 to and wide fractures can prove beneficial.2 pant while it is sheared and heated in sur-
4,000,000 lbm of proppant placed in the High-permeability formation treatments face equipment, well tubulars, perforations
fracture (next page, left ) . are on a far reduced scale. Only a few and the fracture. Otherwise, premature set-
Until recently, treatments were performed pieces of blending and pumping equipment tling of the proppant occurs, jeopardizing
almost exclusively on poor producing wells are required, and pumping times are typi- the treatment. A suite of specially designed
(often to make them economically viable). cally less than one hour, and often only 15 chemical additives imparts important prop-
In the early 1990s, industry focus shifted to minutes. Fluid is pumped at 15 to 20 erties to the fluid. Crosslinkers join polymer
good producers and wells with potential for bbl/min [2.4 to 3.2 m 3/min] with a total chains for greater thickening, fluid-loss
greater financial return. This, in turn, meant volume of 10,000 to 20,000 gal [37.9 to agents reduce the rate of filtration into the
an increased emphasis on stimulating high- 75.7 m3] and total proppant weight of about formation and breakers act to degrade the
permeability formations. 100,000 lbm [45,000 kg] (next page, left ). polymer for removal before the well is
The major constraint on production from This technique has been successful in the placed on production.
such reservoirs is formation damage, fre- North Sea, Middle East, Indonesia, Canada The fracture is created by pumping a
quently remedied by matrix acidizing treat- and Alaska, USA. series of fluid and proppant stages. The first
ments. But acidizing has limitations, and While fracturing treatments vary widely in stage, or pad, initiates and propagates the
fracturing has found an important niche. scale, each requires the successful integra- fracture but does not contain proppant. Sub-
The objective in highly permeable forma- tion of many disciplines and technologies, sequent stages include proppant in increas-
tions is to create short, wide fractures to regardless of reservoir type. Rock mechanics ing concentrations to extend the fracture
reach beyond the damage. This is often experiments on cores, specialized injection and ensure its adequate packing.
accomplished by having the proppant testing and well logs provide data on forma- Fracturing fluid technology has also devel-
bridge, or screen out, at the end, or tip, of tion properties. Sophisticated computer soft- oped in stages. Early work focused on iden-
the fracture early in the treatment. This “tip ware uses these data, along with fluid and tifying which polymers worked best and
screenout” technique is the opposite of what well parameters, to simulate fracture initia- what concentrations gave adequate prop-
is desired in low-permeability formations tion and propagation. These results and eco- pant transport. Then, research on additives
nomic criteria define the optimum treatment to fine-tune fluid properties hit high gear.
design. Process-controlled mixing, blending
and high-pressure pumping units execute
the treatment. Monitoring and recording
devices ensure fluid quality and provide
36 Oilfield Review
Much was learned, but what finally Typically, particulates or other fluid addi-
40 vehicles
emerged was a huge array of complicated tives are used to reduce leakoff by forming a
fluids—difficult to prepare and pump—and filter cake—termed an external cake—on
an amazing assortment of single-use addi- the surface of the fracture face. Acting
tives (most had to be custom manufactured) together with the polymer chains, the fluid-
that required expensive material inventories. loss material blocks the pore throats, effec-
In the past ten years, a more productive tively preventing invasion into the rock
research direction has emerged. Oil compa- matrix (below ) .
1,000,000 gal nies, service companies and polymer manu- This approach has been applied suc-
facturers have concentrated on the basic cessfully for decades to low-permeability
physical and chemical mechanisms underly-
ing the behavior of fracturing fluids in an Flow
attempt to find improved approaches to fluid
External filter cake
design and use. This initiative has led to Bridging zone
major advances, including higher-performing
polymers, simpler fluids, multifunctional
additives and continuous, instead of batch, Invaded zone
mixing. These developments have had a sig-
3 million lbm
nificant, beneficial impact on the industry.
Recent innovations are extending the state Uncontaminated
Propp Propp Propp
formation
of the art in four areas:
Propp Propp Propp • controlling fluid loss to increase
fluid efficiency nZones of leakoff in a fractured formation.
Propp Propp Propp • extending breaker technology to In low-permeability reservoirs, the
improve fracture conductivity invaded zone may be small or nonexis-
tent, since polymer molecules are too
Propp Propp Propp
• reducing polymer concentration to large to enter the matrix. In these cases,
improve fracture conductivity an external filter cake controls fluid loss.
Propp Propp Propp
• eliminating proppant flowback to In high-permeability reservoirs, signifi-
stabilize fractures. cant matrix damage can occur because
Propp Propp Propp
Each provides new opportunities for of particle penetration.
Autumn 1995 37
Measuring Dynamic Fluid Loss
in the Laboratory
Oil-filled annulus
Fluid To differential
slot pressure
transducers Sleeve
Leakoff Core
fluid
Core
holder
Fluid outlet
■Dynamic fluid-loss cell and core holder. A specially designed core holder simulates a slot-flow geometry, the one most representative of the actual fracturing
process. Pressure and leakoff measurements provide the data necessary to evaluate the depth of invasion and the impact of various additives on fluid loss.
The two components of the apparatus are a floating-piston accumulators and coils of tubing
shear history simulator and a fracture simulator. immersed in a temperature-controlled bath.
The first uses a static mixer and 800 ft [244 m] of A computer-controlled valving arrangement
small-diameter tubing to simulate preparation ensures that the fluid always travels in the same
and shearing of the fluid in the well tubulars. direction across the core face.
The second subjects the fluid to shear and temper-
ature conditions of the fracture, using two large,
38 Oilfield Review
350
(< 0.1 md) formations in which polymer
and particulate sizes exceed those of the
pore throats. In high-permeability reservoirs, 300
Calculated
however, fluid constituents may penetrate Laboratory approximation
into the matrix, forming a damaging internal 250
Fracture height = 300 ft
10 md
200
structure degrades with shear, polymer may
8
be able to penetrate the rock matrix.
150 Dynamic tests revealed that commonly
6
used additives were less effective in control-
100 ling fluid loss than static tests had previously
4
Shear rate indicated. Also, a direct link between fluid
50
efficiency and shear rate was demonstrated.
2
The higher the fraction of fluid lost under
high shear early in the treatment, the higher
0
0 the total leakoff volume and the lower the
20 40 60 80 100 120 140
efficiency. Spurt has a dominant effect on
Time, min
efficiency and the volume of fluid pumped,
nEffect of shear rate on dynamic fluid loss. Fluid loss increases with increasing perme- particularly for the short pumping times
ability, illustrating the importance of controlling leakoff in high-permeability forma- encountered in fracturing high-permeability
tions. Maintaining high fluid efficiency is critical to creating a cost-effective fracture formations. If spurt is not controlled quickly
with minimal formation damage. and effectively during the high-shear period,
fluid efficiencies drop dramatically.
These observations prompted researchers
to develop a superior additive system that:
• controls spurt under high shear rates in
high-permeability formations
• minimizes the influence of permeability
on leakoff12
• limits the invasion of polymer into
the matrix
• reduces the overall amount of polymer
pumped into the fracture.
40 Oilfield Review
16
Evaluation of several additive types led to
a combination of particulate materials, the
HIGHSHEAR system, that achieves the
12
Fluid-loss vol, ml
above goals. One agent moves rapidly to
the fracture face during the early stage of No fluid-loss additive
fluid loss when leakoff flux is high. This type
of particle seals a major portion of the 8
exposed surface quickly and adheres 25 lbm/1000 gal best available
securely to the surface, resisting high shear
forces. The second material plugs the 4
remaining gaps in the developing filter cake
as the shear rate drops, significantly reduc- 15 lbm/1000 gal HIGHSHEAR
ing filtrate losses and sealing the surface so
0
that polymer particles cannot penetrate the 10 20 30 40 50
matrix. Laboratory tests show a 25 to 75% Time, min
reduction in spurt compared to the best
available products today (right ). Fluid effi- nEffect of additives on fluid loss. The HIGHSHEAR additive system provides substantial
ciency improves, meaning less fluid to improvement in controlling leakoff compared to the best materials available today.
pump, less fluid to break and easier
cleanup. Since this process is aided by hav- Understanding and Improving Fracture of the residual fluid, thereby allowing more
ing a less viscous fluid—one that promotes Conductivity efficient cleanup. Breakers are used at reser-
high initial leakoff—there is a further oppor- Simply creating a fracture does not guaran- voir temperatures below about 325°F
tunity for cost-savings and improved tee better well performance. The fracture [163°C].16 If breaking is insufficient, con-
cleanup by reducing the polymer concen- must provide a conductive flow path for for- centrated polymer remains in the proppant
tration in the fluid (page 46 ). mation fluids. For decades, poor treatment pack, reducing the conductivity and treat-
One concern with fluid-loss additives has results were blamed on insufficient fracture ment effectiveness.
been the possibility of their presence reduc- length or inadequate proppant transport and Historically, active chemical breakers
ing proppant-pack conductivity. A key ques- placement. Studies in the late 1980s were dissolved in the fluid during surface
tion is: Does the reduction in matrix dam- revealed that substantial fracture damage, mixing. As a result, the fluid was being
age brought about by additives outweigh the and resulting impairment to flow, could be attacked even as it was being pumped. Care
potential damage they may inflict within the caused by polymer residue blocking the had to be taken to use a sufficiently low
fracture itself? As is well-known, the amount pore spaces between proppant particles.15 breaker concentration. Otherwise, viscosity
of conductivity damage is highly dependent Once the proppant pack is placed and would decrease too quickly, and proppant
on the type of fluid used.13 Flowback tests pumping stops, fluid filtrate leaks off into the would settle. With this low breaker concen-
with different fluids and leakoff control rock matrix and the pressure declines. The tration, only partial degradation of the poly-
additives confirm that particulate fluid-loss fluid remaining in the fracture must then be mer occurred. The result was impaired frac-
additives do limit matrix damage by mini- flowed back to allow production of hydro- ture conductivity.
mizing fluid invasion. Conductivity tests carbons. Termed cleanup, this process is crit- Research during the past decade has
show that these agents are actually less ical to the success of the treatment. One way found ways to increase breaking efficiency.
damaging to the proppant pack than previ- to aid cleanup is to ensure that the polymer
ously thought. In most fracturing fluids, their residue has been reduced to a minimum. A 11. Navarrete et al, reference 6.
impact on conductivity is minimal when second is to use as little polymer as possible 12. In high-permeability formations, permeability can
used in low concentrations.14 to start with. The next two sections look at vary widely. Using an average value in job design
calculations can present execution problems during
With these encouraging laboratory results, recent developments in both areas. the treatment. An additive that dampens the effect of
the next step was to test whether the HIGH- permeability variation can be beneficial.
SHEAR system could reduce treatment costs Extending Breaker Technology 13. Brannon HD and Pulsinelli RJ: “Evaluation of the
and improve well productivity better than At the end of a treatment, the fluid left in the Breaker Concentrations Required to Improve the Per-
meability of Proppant Packs Damaged by Hydraulic
conventional products. Extensive field trials fracture has been partially dehydrated due to Fracturing Fluids,” paper SPE 19402, presented at the
in Canada show an overall 15 to 20% cost filtrate loss. The effective polymer concentra- SPE Formation Damage Control Symposium,
Lafayette, Louisiana, USA, February 22-23, 1990.
savings (see “Canadian Treatments Demon- tion can be an order of magnitude higher
14. In instances where the impact is significant, the dam-
strate Cost Savings and Productivity Gains,” than that originally pumped, reaching 300 age can be countered by lowering the polymer con-
next page ). Productivity improvements aver- to 600 lbm/1000 gal [36 to 72 g/cm3]. If the centration since fluids of lower viscosity can be used
aged 460% compared to 260% on offset polymer stays intact, an ultrahigh viscosity, during the treatment.
15. Hawkins GW: “Laboratory Study of Proppant-Pack
wells without the additive. Additional well gelled mass results that blocks the pore Permeability Reduction Caused by Fracturing Fluids
tests are planned in areas such as the US space and cannot easily be flowed back into Concentrated During Closure,” paper SPE 18261,
Gulf Coast where a variety of fluids are used the well. presented at the 63rd SPE Annual Technical Confer-
ence and Exhibition, Houston, Texas, USA, October
and larger-scale, high-permeability fractur- To prevent this, the polymer is attacked by 2-5, 1988.
ing treatments are performed. fluid breakers—oxidizers or enzymes that 16. Above this level, the fluid breaks with time due to the
sever the polymer chain at its weakest thermal degradation of the guar molecules.
points, degrading it into smaller, more
mobile fragments. This reduces the viscosity
Autumn 1995 41
Canadian Treatments Demonstrate
Cost Savings and Productivity Gains
permeability variations.
20
Pad volumes of 2000 to 2500 gal [7.6 to 9.5m3]
were used with total slurry volumes of 4000 to
15
6000 gal [15.1 to 22.7 m3] in four stages. Typical
rates were 15.7 bbl/min [2.5 m3/min]. Maximum
sand concentrations ranged from 16.5 to more 10
42 Oilfield Review
Fractional retained permeability
By encapsulating breakers so they do not 1
fluid pumped meant less fluid to break. Cleanup and much controversy. Recent findings have Fractional reduction in porosity
Retained permeability, %
25
test results and reproducibility were also optimal breaker for a given temperature
quantified, leading to a refined testing pro- range and development of an additive to
20
cedure. Testing by these laboratories using assist in removal of the degraded polymer.
the new guidelines demonstrates that reli- Each type and composition of breaker are 15
able results can be achieved, provided test effective over a certain limited temperature
conditions are closely controlled. Testing range, based on performance and cost crite- 10
within the same laboratory is expected to ria. This requires tailoring of chemical prop-
vary by 15 to 20% while lab-to-lab varia- erties and encapsulation technology so that 5
tions have a 25 to 30% relative deviation the breaker remains active in the proppant
( right ) . Before uniform procedures, test pack long enough to do its job. A suite of 0
results often deviated by 100% or more. materials is needed to cover the range of Lab 1 Lab 2 Lab 3
temperatures that are encountered in frac-
turing operations.
nInterlaboratory comparison of conduc-
100
tivity results. Major improvements in
Borate crosslinked guar But a breaker by itself is often not enough. reproducibility stem from consistency in
Retained permeability, %
Temperature = 160°F Conductivity testing shows that residual testing methods and close control over
80
fragments left after the primary polymer test conditions, removing a major road-
block in correlating data between labs.
60 structure has been degraded can still cause
significant pore blockage (below ) . Worse,
40 under certain conditions, the fragments can flow paths. Pore blockage is reduced and
coagulate—or bind together into a viscous proppant-pack permeability increases.
20 mass—and reduce conductivity further. Laboratory testing in standard conductiv-
Sometimes, adding more breaker can aggra- ity cells shows that the breaker plus addi-
0
vate coagulation. tive system outperforms breaker alone,
0 100 200 300 To avoid this, a special blend of anticoag- yielding almost 40% higher proppant-pack
Polymer concentration, lbm/1000 gal
ulants was developed. This blend, the permeabilities at low-end additive concen-
nEffect of polymer concentration on CleanFLOW additive, works synergistically trations (next page, top). Retained perme-
retained permeability. As the residual with the breaker to reduce the size of the abilities improve to as much as 90% at
polymer concentration remaining in the polymer fragments and prevent their ten- increased concentrations.
proppant pack increases, the permeability dency to coagulate. The dispersive action of In field testing, treatments with the
decreases, emphasizing the need for com- the additive increases mobility and available CleanFLOW additive showed higher levels
plete polymer degradation and removal.
of polymer returns during flowback than
previous treatments on offset wells. Higher
returns are a direct measure of improved
proppant-pack cleanup. Over a given flow-
back period in a well in western Wyoming,
+ breaker
USA, 62% of the polymer was returned
when the additive was used with breaker,
compared to 31% for an offset well with the
breaker alone. At the end of the test phase,
the well treated with the CleanFLOW addi-
+ CleanFLOW
and breaker tive was still returning polymer, while the
offset well was not (next page, bottom ).
Net present value (NPV) calculations on
various formation types show that conduc-
tivity-related production increases realized
nEnhanced proppant-pack cleanup. Use of a breaker alone (top) from the use of the additive can significantly
results in polymer fragments that may be difficult to remove improve well economics.
from the fracture. The CleanFLOW additive system works with
the breaker (bottom) to prevent fragment coagulation, facilitate
cleanup and improve fracture conductivity.
44 Oilfield Review
Reducing Polymer Concentration vated temperatures, however, higher viscosi- els can be reduced to about 25 lbm/1000
Today, over 70% of the fracturing treatments ties have historically been used. In non- gal [3 g/cm3]. The goal of recent research
conducted use guar- or hydroxypropyl guar- crosslinked (linear) systems, this means has been to formulate a reliable fluid using
based fluids.18 The rheology of such fluids adding more polymer, resulting in polymer even less polymer to reduce cost and
has been studied for years. When added to concentrations of 40 lbm/1000 gal [4.8 improve fracture conductivity.
water, guar molecules hydrate and swell, g/cm3] or higher. This approach is expensive Various metal ions, including titanium
increasing in diameter and length. The and often results in fluid mixing and han- and zirconium, have been used for decades
hydrated strands overlap and hinder motion, dling difficulties. Fracture conductivity can as crosslinking agents. In recent years,
giving rise to an increased viscosity of the be impaired, since more polymer must be boron [as B(OH)4-] has grown in popularity
solution. For adequate proppant transport broken and produced back. Crosslinking and is by far the most common element
and placement, a viscosity of 100 centipoise has become a common means of enhancing today. Various boron salts and compounds
(cp) at a shear rate of 100 sec-1 is generally viscosity at lower polymer levels. While
accepted as a minimum guideline. effective in building viscosity, this practice 18. Gulbis J: “Fracturing Fluid Chemistry,” in Econo-
mides MJ and Nolte KG (eds): Reservoir Stimulation,
To minimize fluid leakoff and counteract can lead to complicated, hard-to-break 2nd ed. Englewood Cliffs, New Jersey, USA: Prentice
the inherent thinning of guar systems at ele- structures. Using crosslinkers, polymer lev- Hall (1989): 4-1–4-14.
40 nConductivity
comparison. The
Temperature = 225°F
new additive sys-
tem provides sub-
30 stantially greater
Retained permeability, %
retained perme-
abilities compared
with no breaker or
20 with use of a
breaker alone, con-
firming that the
additive enhances
10
removal of residual
polymer and limits
pore blockages.
0
No breaker Breaker only CleanFLOW CleanFLOW
system system
2 gal/1000 gal 4 gal/1000 gal
40
western Wyoming,
30
USA, 62% of the
polymer was
25 returned during
62% polymer return flowback. For an
20 offset well with
breaker alone, dur-
15 ing a similar flow-
back time, only
10
31% of the polymer
was returned.
31% polymer return
5
0
200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Cumulative flowback volume, bbl
Autumn 1995 45
can be used (below ) . If the pH19 of the fluid ing extended pumping times. For these 1
75°F
is above 8, crosslinking occurs almost applications, it is necessary to delay the
Galactose
nStructure of guar other criteria must be satisfied: the chains
CH2OH CH2OH and crosslinking need to have enough active crosslink sites
substituents
O O by boron. Guar is and the proper number of borate ions must
HO H HO H composed of a
H H be present to build a network structure.
mannose back-
Careful balance between the two has to be
H OH H
H
OH H
Mannose bone with intermit-
backbone tent galactose sub- maintained to produce a stable system.
O O stituents (top). In the past, borate crosslinked fluids at the
H OH H OH
Boron, in the form lower threshold of 20 lbm/1000 gal guar
CH2 CH2 CH2OH CH2 of B(OH)4-, reacts concentration have been formulated in the
O O O O with the hydroxyl
H H H H H (OH) groups on the laboratory and even tested in the field.
H H
O O O O polymer in a two- Operational experience, however, showed
OH OH H OH OH H OH OH H OH OH H step process to link them to be unreliable and overly sensitive to
two polymer small variations in fluid chemistry.
strands together
H H H H H H H H
(bottom).
How can the threshold polymer level nec-
essary for crosslinking be reduced success-
Borate Crosslinking of Guar fully? Based on earlier work with zirconate
systems, scientists evaluated what types of
materials associate with guar molecules to
B(OH)3 + OH- B(OH)4- increase their solubility. After studying the
behavior of a variety of materials, researchers
identified a combination of chemicals that
OH O OH
permits stable borate crosslinking at polymer
B(OH)4- + R R B + 2H2O
OH
concentrations as low as 15 lbm/1000 gal
O OH
and reliable field formulations to be mixed in
the 15- to 20 lbm/1000 gal range (next page,
OH O O
bottom ).
B(OH)4- + 2R R B R + 4H2O These systems, referred to as low-guar flu-
OH O O
ids, can be used at fluid temperatures up to
R= Mannose backbone 175°F [80°C]. They exhibit viscosities nor-
mally measured for borate crosslinked fluids
with polymer concentrations 5 to 10
lbm/1000 gal higher.
Proppant suspension properties are excel-
lent, allowing use of proppant concentra-
tions as high as 17 to 20 pounds of proppant
added (ppa). The fluid can be continuously
mixed and then crosslinked. It can be effec-
tively degraded using available breaker sys-
46 Oilfield Review
tems for rapid, efficient cleanup following Permeability Percent Retained nRetained conduc-
Fluid Temperature (°F) (darcies) Conductivity tivity of low-guar
the treatment. Conductivity tests on proppant systems. By reduc-
packs placed with low-guar fluids show 150 2% KCI 100 244 -
ing the polymer
15 lbm 100 148 61
to 200% higher pack permeabilities than concentration,
20 lbm 100 103 42 cleanup is easier
those obtained with conventional polymer 2% KCI 125 -
216 and the retained
concentrations (right ) . 15 lbm 125 130 60 conductivity of the
Over 500 treatments have been pumped 20 lbm 125 106 49 proppant pack is
since the introduction of the fluid in 1994. 30 lbm 125 63 29 greater than for
Because of the success of these treatments, 2% KCI 175 125 - conventional fluids
the low-guar system is rapidly becoming the 20 lbm 175 61 49 that require higher
30 lbm 175 32 26 polymer concen-
fluid of choice for applications to 175°F. trations.
Reduced polymer loading can lead to Proppant Type: Sand
increased fluid leakoff. This can be averted Proppant Concentration: 2 lbm/gal
by the adding fluid-loss agents, particularly
in higher-permeability formations. Field stud- Controlling Proppant Flowback
ies, however, have shown that this may not Flow of proppant into the wellbore follow- widths and higher proppant concentrations
always be necessary, depending on the type ing a fracturing treatment is of major con- have become the norm. In areas such as
of formation and its permeability. In more cern. This phenomenon may occur during Alaska and the North Sea, up to 20% of the
than 130 jobs performed in Kansas, USA in initial cleanup or sometime after the well is proppant may be produced back, while
5- to 50-md formations, no fluid-loss addi- put back on full production. Termed prop- some instances of up to 50% have been
tives were used. Fluid efficiency was not pant flowback or proppant backproduction, reported.20 This can translate into anywhere
adversely affected, and conductive fractures it can lead to expensive, time-consuming from 1000 to 100,000 lbm [454 to 45,400
were obtained. Cleanup times were reduced remedial operations and safety concerns. In kg] of proppant per treatment. Although the
by 50% compared to previous treatments. low-rate wells, proppant may settle in the flowback may stop with time, many wells
Proppant placement efficiencies—the per- casing, requiring periodic wellbore produce proppant throughout their life-
centage of proppant effectively placed in the cleanouts. Loss of near-wellbore fracture times. Wells often must be placed on
fracture—met or exceeded 97%. conductivity can result, and production may restricted chokes to limit pressure drops
In even higher-permeability formations, cease entirely if the productive zone is fully and stabilize the proppant pack.21
as tests in Alberta, Canada (page 42 ) con- covered. In high-rate wells, erosion occurs While changes in fracture design and exe-
clusively demonstrate, low-guar systems to tubulars, control valves and wellhead cution can sometimes alleviate the problem,
can be used effectively with the new equipment. Disposal costs for produced a typical solution has been the use of resin-
HIGHSHEAR fluid-loss additive to improve proppant may be substantial. coated proppant (RCP). RCPs are pumped
fracture conductivity, well cleanup and well The frequency of this problem has into the fracture near the end of the treat-
performance. increased markedly as greater fracture ment, referred to as tailing in. The well may
15 lbm/1000 gal 20 lbm/1000 gal 19. The measure of acid intensity equal to the logarithm
180 400
nViscosity of low- of the reciprocal of the hydrogen ion concentration
of the solution. pH 7 is neutral; below 7 is acidic,
Viscosity at 100 sec-1, cp
Viscosity at 100 sec-1, cp
Autumn 1995 47
nPropNET proppant the proppant particles, providing increased
flowback control.
Random glass strength, or that they stabilize and distribute
fibers pumped in stress, aiding bridging, within a significant
the fracturing fluid area of the pack. The fiber structure is more
form a net-like flexible than cured RCPs, allowing the prop-
structure that stabi-
lizes the proppant pant-fiber pack to shift without failing.
pack while allowing Laboratory tests show that the ability of
high production the pack to resist proppant flowback is a
rates of formation function of fiber concentration. Stability
hydrocarbons. increases with fiber content until a plateau
is reached (below ). While laboratory data
show that use of 1.5% fibers by weight can
reduce permeability by up to 30% com-
pared to packs without fibers,23 field results
show less reduction. Conductivity values for
be shut in for a period of time to allow the material forms a web, or network, which packs with fibers are superior to those mea-
resin to cure as its temperature rises, binding stabilizes the proppant-fiber pack and sured for postcured RCPs.24
the proppant particles together at their allows high production rates of oil or gas Despite the low concentration of 1.5%,
points of contact. Ideally, a consolidated, (above ) . The technology is based on the the fiber level is about 30% by number of
high-conductivity matrix is formed. principles of fiber reinforcement commonly particulates, or about one fiber for every two
RCPs, however, are not universally appli- used in a variety of industrial and commer- proppant particles. Fiber length is an order
cable and have certain severe limitations. cial applications as a strengthening method. of magnitude more than proppant diameter,
Performance is sensitive to shear, tempera- For example, natural and synthetic fibers are so, for example with sand proppants, each
ture, closure pressure and shut-in time. Con- used to protect dams and other concrete fiber touches approximately five particles
ductivities are frequently lower than and soil structures from erosion. The inher- (next page, top).
expected. In low-temperature wells, an ent ability of fibers to stabilize highly Single-phase flow tests using water and
expensive activator must be added to the porous, particulate-containing materials two-phase tests with water and gas were
RCP at a 0.5 to 2.0% concentration for the provided a basis for these investigations.
resin to cure. Shut-in times may be as long A comprehensive set of tests has been
as 24 hours. Resin coatings can interact applied to determine the applicability of 9
48 Oilfield Review
nPhotomicrograph
of fiber-reinforced
proppant pack.
The glass fibers are
about an order of
magnitude longer
than the diameter
of a typical prop-
pant particle,
allowing each fiber
to contact about
five particles. The
fibers stabilize the
pack by inter-
weaving among
particles, provid-
ing increased
strength and sta-
bility. They may
also promote
bridging and stress
distribution within
the pack.
gradients in excess of 100 psi/ft, but typical the type of minerals present in the water (sil-
140
guidelines. Under these conditions, pack 300°F [149°C] when in contact with a sil- 80
stability reaches a minimum at about 40% ica-saturated formation brine. Life 60
of the maximum pack strength in single- expectancy could be greater, depending on 40
phase fluid flow. Based on laboratory data, downhole conditions.
20
maximum cleanup rates of 30 bbl/day/per- Glass fibers do not interact with common
0
foration [4.8 m3/day/perforation] for sand fracturing fluid systems or additives, a key
0 20 40 60 80 100
and 20 bbl/day/perforation [3.2 m3/day/per- concern with RCPs. Their presence in the Nitrogen content, % total flowing
foration] for ceramic proppants are being fluid slurry also reduces proppant settling,
used in the field. aiding proppant transport and placement. nResistance of sand-fiber packs to prop-
Tests results show that no minimum con- Glass fibers have certain limitations to be pant flowback. At a fiber content of 1.5%,
the pressure drop the pack can withstand
fining stress, shut-in time or reservoir tem- considered during treatment design. They is a maximum when no nitrogen is pre-
perature are required for the fibers to be are not effective at temperatures above sent, exceeding 100 psi/ft. As the nitro-
effective, overcoming some of the severest 300°F and under certain conditions: if for- gen content increases, pack strength
limitations of RCPs. decreases until a minimum is reached at
Packs were cycled under stress to simulate about an 80% gas volume.
shut-in and production periods, with packs 22. Almond SW, Penny GS and Conway MW: “Factors
Affecting Proppant Flowback with Resin Coated
subjected to alternating stress levels of 1000 Proppants,” paper SPE 30096, presented at the Euro-
psi and 4000 psi. No pack failures were pean Formation Damage Conference, The Hague,
The Netherlands, May 15-16, 1995.
observed, even after more than 30 cycles, 23. This observation is consistent with calculations in
for both sand and ceramic proppant packs. which fibers represent 5 to 6% of the pore volume.
Aging studies were also conducted, since 24. Card RJ, Howard PR and Feraud J-P: “A Novel Tech-
nology to Control Proppant Backproduction,” paper
glass fibers can be dissolved by formation SPE 31007, SPE Production & Facilities (in press).
waters. The solubility rate depends on sev-
eral factors, including temperature, pH and
Autumn 1995 49
Economic Benefits of Fiber Reinforcement: Field Results
More than 150 hydraulic fracturing treatments using PropNET fiber-reinforced packs have been conducted in
the USA (top) and Venezuela. Here is a look at three typical examples.
Texas, USA:
A 10,000-ft [3080-m] gas well in south Texas,
USA with a bottomhole temperature of 275°F
[135°C] was fractured using a borate-crosslinked
fluid. A fiber-reinforced pack was placed with
1.5% fiber by weight of proppant. Fibers were
added during the entire proppant stage. Pumping
pressure levels were similar to treatments with-
out fibers present. Of the nearly 16,000 lbm
[7300 kg] of proppant placed, less than 0.07%
■Fracture treatment in south Texas.
was produced back. The well was later cycled
through four shut-in and production periods, with
a shut-in closure stress of 1900 psi and a flowing ■ Fluid returns in a
400 PropNET south Texas well.
tubing pressure of 4200 psi. Production was
Faster flowback rates
essentially proppant free and remains so today and larger cumulative
Cumulative fluid returned, bbl
well over a year later. Productivity from the well flowback volumes
300
exceeded that from offsets. were found with the
PropNET system than
In another south Texas gas well, having two Gas
with RCPs. Gas break-
sandstone layers separated by a shale layer, through was much
200
226,000 lbm [102,500 kg] of proppant were Gas
RCP earlier and product
placed using a borate-crosslinked guar fluid. sales were realized
sooner.
Here, fibers were tailed in with the last 15% of 100
the proppant. Flowback started as soon as the
treatment was completed. The initial flowback
rate of 500 bbl/day [80 m3/day] was increased
0
10 20 30 40 50 60 70
to 1000 bbl/day [160 m3/day]. Less than 0.05%
Flowback time, hr
of the proppant was produced back over a
four-day period.
30,000
How fiber reinforcement compares with RCPs ■ Decrease in flowback
was evaluated in an offset well. An upper produc- costs for a south Texas
well. With a much
tive zone was fractured with a 15% fiber tail-in
shortened flowback
Flowback costs, $
50 Oilfield Review
mation water is not silica-saturated, if con- New Solutions with Novel Technology
fining stress exceeds the crush strength of The impressive developments in fracturing
the proppant, or if the fracture will be fluids technology from 1985 to 1993 have
treated later with certain types of acid.25 been reinforced by innovations during the
While fibers could be distributed through- past two years. Advanced fluid-loss addi-
out the entire length of the proppant pack, tives are improving fluid efficiency at lower
field experience supports laboratory tests fluid viscosities. Combined with new low-
showing that only the tail-in portion of the guar systems, this means reduced costs and
According to Gary Slusher, project production treatment, typically the last 15 to 20% of the reduced damage to the fracture. Enhanced
engineer for Enron Oil & Gas, in Corpus Christi, proppant pumped, normally requires fiber breaker and additive systems are speeding
Texas, “We have had to live with RCPs. They can addition. If, however, there are concerns well cleanup, giving more complete fluid
be effective, but they are expensive and require
that the entire height of the producing zone degradation and cleaner, more conductive
cannot be covered or that sand control will fractures. And, the fracture is now more sta-
long cure times, and once you start pumping
be a problem, it is prudent to use fibers ble, thanks to innovations in proppant flow-
them you’re committed. When we bring a well throughout the proppant pack. back control.
back, we have to do it gingerly to avoid prob- Extensive field experience in the USA Individually and collectively, these new
lems. With PropNET, we can start and stop pump- (see “Economic Benefits of Fiber Reinforce- technologies are benefiting oil and gas oper-
ing as necessary and aggressively flow back the ment: Field Results,” previous page ) has ators by reducing treatment and well
demonstrated the superior performance of cleanup costs, increasing well productivity
wells. The system is more effective in stopping
fiber-reinforced packs, including reduced and speeding product sales to market. In the
proppant production than RCPs. shut-in times, faster cleanup rates and future, the increased emphasis on more effi-
“We’ve seen both direct and indirect benefits greater execution efficiency. This translates cient and effective fluids and the synergistic
with PropNET. We’re getting much higher fluid directly into rapid well turnaround and application of low-cost innovations will
recoveries and more rapid cleanup. This reduces decreased wellsite costs. yield further economic gains. —DEO
our total job costs substantially, by 12 to 15% or
On average, less than 0.2% of the prop-
pant pumped was produced back in over 25. Fibers are resistant to hydrochloric acid [HCl], the
more. We believe our effective frac lengths are
150 treatments. Cleanup rates as high 4000 acid commonly used, but not to hydrofluoric acid
longer, too. But, most importantly, we get the gas bbl/day [636 m3/day] water and gas rates of [HF] which is known to dissolve glass.
to market sooner by turning the wells around up to 10 MMscf/day were observed, 10
faster, sometimes in about 24 hours instead of times those used by local operators for
seven to ten days. We’ve also been able to use RCPs. Since cleanup can represent 10 to
60% of the total cost of a fracturing treat-
simpler fluids with less additives. That’s saving
ment, this time savings has a dramatic eco-
us additional money and reducing friction pres- nomic impact.
sures and horsepower requirements. When the With the concept proven both in the labo-
job has been executed according to design and ratory and field, this technology will rapidly
the material has been placed where it’s needed, be applied beyond the Western Hemi-
the results have been excellent.”
sphere. To support this expansion, alterna-
tive fibers are being tested for application
above 300°F in areas such as the North Sea.
Indiana, USA:
In low-temperature gas wells in Indiana, USA,
RCPs are commonly used when fracturing multi-
ple zones. This requires an activator and an
extended shut-in period for curing of each zone
after treatment, or a total time of four to five days
per well. When fiber technology was used
instead, the initial zone could be flowed back
only 10 min after the job was completed. Follow-
ing a limited flowback period, the next zone could
be fractured immediately, completing operations
in one day, a 75 to 80% savings in rig labor, rig
time and equipment costs.
Autumn 1995 51